Stable Isotope Ecology of Land Snails from a High-Latitude Site Near Fairbanks, Interior Alaska, USA

Quaternary Research 83 (2015) 588–595

Contents lists available at ScienceDirect

Quaternary Research

journal homepage: www.elsevier.com/locate/yqres

Stable isotope ecology of land snails from a high-latitude site near Fairbanks, interior Alaska, USA

Yurena Yanes

Department of Geology, University of Cincinnati, Cincinnati, OH 45221, United States article info abstract

Article history: Land snails have been investigated isotopically in tropical islands and mid-latitude continental settings, Received 2 June 2014 while high-latitude locales, where snails grow only during the summer, have been overlooked. This study pre- Available online 7 April 2015 sents the first isotopic baseline of live snails from Fairbanks, Alaska (64°51′N), a proxy calibration necessary prior to paleoenvironmental inferences using fossils. δ13C values of the shell (−10.4 ± 0.4‰) and the body Keywords: (−25.5 ± 1.0‰) indicate that snails consumed fresh and decayed C -plants and fungi. A flux-balance mixing Land snails 3 Stable isotopes model suggests that specimens differed in metabolic rates, which may complicate paleovegetation inferences. δ18 − ‰ ‰ δ18 fi Boreal forest Shell Ovalues( 10.8 ± 0.4 )were~4 higher than local summer rain O. If calci cation occurred during 18 Fairbanks summer, a flux-balance mixing model suggests that snails grew at temperatures of ~13°C, rainwater δ Ovalues Quaternary of ~−15‰ and relative humidity of ~93%. Results from Fairbanks were compared to shells from San Salvador (Bahamas), at 24°51′N. Average (annual) δ18O values of shells and rainwater samples from The Bahamas were both ~10‰ 18O-enriched with respect to seasonal (summer) Alaskan samples. At a coarse latitudinal scale, shell δ18O values overwhelmingly record the signature of the rainfall during snail active periods. While tropical snails record annual average environmental information, high-latitude specimens only trace summer season climatic data. © 2015 University of Washington. Published by Elsevier Inc. All rights reserved.

Introduction Laboratory experiments by Stott (2002) and Metref et al. (2003) indicated that the δ13C values of the shell are a function of the δ13C Land snails contain an aragonitic shell that is fairly durable in values of the consumed and assimilated plant matter, whereas archeological and paleontological sites and, accordingly, is commonly atmospheric CO2 and carbonate ingestion had a minor effect in the preserved in Quaternary outcrops (Goodfriend, 1992, 1999). Common- shell δ13C values. However, this does not seem to be the case in every ly, land snails are the only biological material that is both abundant and setting and for all taxa because snails occupying carbonate-rich well-preserved in terrestrial sedimentary records, so they may be the areas may incorporate carbon from limestone ingestion into their only biotic source of paleoclimatic information. In particular, the carbon shells (Goodfriend and Hood, 1983; Goodfriend, 1987; Yanes et al., (δ13C) and oxygen (δ18O) isotopic composition of land snail shelly re- 2008; Pigati et al., 2010, 2013). Moreover, a recently published laborato- cords are valuable terrestrial proxies generally used to reconstruct the ry experiment by Zang et al. (2014) proposed that although the snail paleovegetation (e.g., relative abundance of photosynthetic pathways) shell is predominantly influenced by diet, both carbonate ingestion and the paleoatmospheric conditions (e.g., rainfall and humidity), and atmospheric CO2 could also play a role. Balakrishnan and Yapp respectively (Balakrishnan and Yapp, 2004). However, the calibration (2004) pointed out that different individuals of the same and different and validation of land snails as a credible environmental proxy are not taxa might vary in metabolic rates, which could be reflected in the as straightforward because both the δ13C and the δ18Ovaluesareinflu- snail carbon isotope pool and may further complicate paleovegetation enced by multiple variables simultaneously (Balakrishnan and Yapp, inferences. 2004), and the magnitude at which various physical parameters affect While no published laboratory experiment has attempted to moni- the isotopic signature of the shell may vary between species, localities tor the controlling factors of shell δ18Ovaluestodate,Balakrishnan and even the temporal and spatial scale considered. Thus, studies that and Yapp (2004) used empirical and theoretical data and concluded quantify the environmental significance of the isotopic signature of that the dominant variables controlling the δ18O values of the shell inland snails using present-day samples are timely and highly valuable clude rainwater δ18O values, water vapor δ18O values, relative humidity for paleoclimatologists and paleontologists. and temperature. It is generally assumed that O2 from the atmosphere and the δ18O values of water derived from ingested plants both have a negligible effect in the snail oxygen isotope pool (Balakrishnan and E-mail address: [email protected]. Yapp, 2004). Despite the large number of environmental variables

http://dx.doi.org/10.1016/j.yqres.2015.03.004 0033-5894/© 2015 University of Washington. Published by Elsevier Inc. All rights reserved. Y. Yanes / Quaternary Research 83 (2015) 588–595 589 controlling snail shell δ18O values, the majority of published field stud- Methods ies suggest that snails seem to be principally influenced by rain δ18O values (Lécolle, 1985; Zanchetta et al., 2005; Yanes et al., 2008, 2009) Present climate in Fairbanks or by both rain δ18O values and relative humidity (Balakrishnan et al., 2005a,b; Yanes et al., 2011a). Conclusively, the environmental meaning Fairbanks, interior Alaska (Fig. 1A), exhibits a continental climate, of the isotopic codes of land snail shells can be complex to understand with extreme seasonal variability in solar radiation. The Alaskan but potentially useful to deduce informative aspects of past continental Climate Research Center (http://climate.gi.alaska.edu/) indicates that climates. for the recording period between 1981 and 2010, annual precipitation Apart from the relative temporal continuity of fossil shells in averaged 274.6 mm, ranging from 54.9 mm in July to 6.4 in March Quaternary sedimentary records, land snails exhibit a large spatial pres- (Fig. 2A). The mean annual air temperature is −2.5°C, varying from ence along latitude, ranging from the tropics to the high-arctic tundra 16.9°C in July to −22.2°C in January (Fig. 2A). Snow falls year-round ex- (Pearce and Örstan, 2006). However, this proxy has been principally cept for the summer months (Fig. 2B). Average annual relative humidity calibrated and used in tropical-subtropical oceanic islands (Baldini is 61.8%, and fluctuates from 72.5% in December to 48.5% in May et al., 2007; Yanes et al., 2008, 2009, 2011a, 2013a; Yanes and (Fig. 2B). Maximum relative humidity values reach up to ~99% in Romanek, 2013) and mid-latitude coastal and inland continental sites August. The climate in Fairbanks during the summer (from May (e.g., Yapp, 1979; Lécolle, 1985; Goodfriend and Ellis, 2000, 2002; to September), when snails are active and grow their shell, is character- Zanchetta et al., 2005; Balakrishnan et al., 2005a,b; Kehrwald et al., ized by mean air temperatures of ~13°C, total precipitation of ~181 mm, 2010; Colonese et al., 2007, 2010a,b, 2011, 2013a,b; Yanes et al., and average relative humidity of ~57% (Figs. 2A–B). The stable isotope 2011b, 2012, 2013b; Yanes et al., 2014). composition of the summer precipitation in Fairbanks is available The present work investigates the isotopic composition of land from the Bonanza Creek ELTER webpage (http://www.lter.uaf.edu/ snails from a woodland ecosystem in Fairbanks (Alaska), at the data_detail.cfm?datafile_pkey=66). For the recording period between latitude of 64°51′N, longitude of 147°49′W, and elevation of 189 m June 2009 and August 2010 (snail samples in this study were live col- (a.s.l.). This study presents land snail stable isotope systematics from lected in August 2010), the summer rainwater exhibited an average the highest-latitude continental interior locality reported in the δ18Ovalueof−15 ± 3‰ (n = 204), ranging from −6.5‰ to −23.6‰ published literature after Yapp (1979), who analyzed four land snail (Figs. 2C–D). shells from the coastal site of Sandnessjoen (Norway), at a latitude of ~66°N. Quaternary permafrost soils and eolian deposits from Hibernation of land snails Alaska are rich in fossil land snails (Pigati et al., 2013) and these materials would potentially provide valuable insights into past climates Land snails are ectothermic, i.e., they do not regulate their internal at high latitudes. But before ancient shells from Alaska can be used body temperature. Consequently, snails that inhabit high-latitude re- to deduce paleoenvironments, proxy calibration and validation are gions, which receive substantial amounts of snow during part of the necessary using present-day samples. This work discusses the environ- year and below zero air temperatures during many months, hibernate mental significance of carbon and oxygen stable isotope values of for a large number of months (Gomot de Vaufleury, 2001). Snails can modern land snail tissues (body and shell) and potential dietary sources be either freezing-tolerant (those who survive freezing body fluids) or (vascular and non-vascular plants, soil organics and fungi) from a freezing-intolerant (those who cannot stand freezing body fluids and boreal forest of Fairbanks and explores the utility of land snails as an survive by extending their supercooling ability) (Ansart et al., 2002). environmental archive at high latitudes. Isotopic data from Alaska Snail physiological and ethological mechanisms to survive cold hardi- (64°N) are explored using published flux-balance mixing models by ness include (Nicolai et al., 2010, 2012): (1) reducing the volume of Balakrishnan and Yapp (2004). Finally, modern land snail samples freezable water, (2) accumulating cryoprotectans (e.g., polyhydric from the tropical island of San Salvador, Bahamas (at the latitude of alcohols, saccharides, free amino acids and lipids) by body water loss ~24°N) were analyzed and discussed for comparison with Alaskan and increasing of body fluid osmolarity, which lowers the freezing counterparts. point of the snail body fluid, (3) borrowing into the soil and (4) forming

Figure 1. Geographical location of the two sampling sites (Fairbanks and San Salvador Island) in this study (A). Photographs of the interior boreal forest sampled (B) and the snail species (Succinea strigata) collected (C) in Fairbanks. Photographs of the coastal environment sampled (D) and the snail species (Succinea barbadensis) gathered (E) in San Salvador, Bahamas. 590 Y. Yanes / Quaternary Research 83 (2015) 588–595

Figure 2. Present climate in Fairbanks. (A) Mean monthly temperature and precipitation. (B) Average monthly relative humidity and snowfall. (C) Deuterium and oxygen stable isotope values of summer rainfall in Fairbanks. Plotted data come from June, July, August and September 2009 and May 2010. (D) Oxygen stable isotope values of the summer rainfall in Fairbanks (from June 2009 to May 2010) by month.

an epiphragm. Many land snail species from high-latitudes are, in part, San Salvador, Bahamas freezing tolerant, i.e., they can survive some ice formation within their body for a limited time, with variable supercooling abilities (Ansart For comparative purposes, a total of n = 15 Succinea barbadensis et al., 2002). Snail mortality during hibernation can be high and may specimens were collected in the tropical carbonate-rich island of San control community dynamics, however, despite relatively high mortal- Salvador (Bahamas) during July 2010 (Table 2). Fresh dead shells of ity rates, high-latitude snails may live from several months to more than Succinea were gathered from a coastal site near Cockburn Town. The 2yr(Nicolai et al., 2012). sampling locality was characterized by open-vegetation dominated by shrubs and grasses. The mean annual temperature in this island is ~24°C and the amount-weighted rain δ18O value is −4.5‰ (SMOW), Target land snail species and sampling protocol on average. Organic matter samples from this locality were not available. For additional details about the geology and environmental setting Fairbanks of San Salvador see Yanes and Romanek (2013). All snail samples (n = 35 total) were live-collected in the woodlands of Fairbanks, interior Alaska, about 1 km north of the University of Organic matter sample collection and preparation Alaska, Fairbanks, during August 2010 (Table 1). After visual searching for snails, two species of small (b5 mm) land snails were found: Several potential carbon food resources of snails, including fresh tree Succinea aff. strigata (n = 25) and Euconulus aff. fulvus (n =10).Several leaves (n = 1), mosses (n = 1), fungi (n = 1) and soil organic matter specimens of each species were deposited in the Carnegie Museum of (n = 1) were collected next to the sampled snail assemblage in Natural History at Pittsburg, with catalog numbers of CM139308 for Fairbanks (Table 3;Fig. 1B). In addition, snail body tissues of some Euconulus and CM139309 for Succinea. The family Succineidae exhibits Succinea specimens (n =20outof25Succinea individuals) and homog- a cosmopolitan geographical distribution, being present in almost enized snail feces from several Succinea specimens (n = 1) were select- every continent, and exhibits an extensive fossil record, especially in ed for isotopic analysis. About 1.5 mg of snail tissue and ~5 mg of plant Quaternary loess deposits of North America (Pigati et al., 2010, 2013) and fungi tissue were weighed in a tin capsule, crimped and combusted and Europe (Moine et al., 2005, 2008). Succineid species often live in an Elemental Analyzer (EA). The CO2 and N2 produced after combus- from 1 to 2 yr and are associated with swamp and periodically ﬂooded tion were analyzed using the IRMS. Analytical uncertainty was ±0.1‰. areas. Generally, succineids have a semelparous life cycle, reproducing in the summer and hibernating during winter (Örstan, 2010). The fam- Carbonate sample preparation ily Euconulidae also exhibits a widespread geographical distribution. In particular, Euconulus fulvus is broadly present in North America, espe- Shells were cleaned in distilled water and ultrasonication, and dried cially in cool and humid soils containing dead wood, and seems to toler- at 40°C overnight. Entire shells (n = 35 total) were ﬁnely ground man- ate non-calcareous soils. ually using an agate mortar and pestle. Entire shell was preferred over Y. Yanes / Quaternary Research 83 (2015) 588–595 591

Table 1 Stable carbon and oxygen isotope results of live-collected land snails from Fairbanks (Alaska) collected during August 2010.

Snail ID Species Locality Latitude Body δ13C Shell δ13C Δ13C Shell δ18O (PDB) (PDB) (Shell-Body) (PDB)

FAI-snail-1 Succinea strigata Fairbanks, Alaska 64°51′N −10.9 −10.3 FAI-snail-2 Succinea strigata Fairbanks, Alaska 64°51′N −10.3 −10.4 FAI-snail-3 Succinea strigata Fairbanks, Alaska 64°51′N −10.5 −11.3 FAI-snail-4 Succinea strigata Fairbanks, Alaska 64°51′N −10.4 −10.8 FAI-snail-5 Succinea strigata Fairbanks, Alaska 64°51′N −24.9 −10.2 14.7 −10.9 FAI-snail-6 Succinea strigata Fairbanks, Alaska 64°51′N −25.3 −10.4 14.9 −10.7 FAI-snail-7 Succinea strigata Fairbanks, Alaska 64°51′N −26.1 −10.8 15.3 −10.7 FAI-snail-8 Succinea strigata Fairbanks, Alaska 64°51′N −24.2 −10.2 14.0 −10.8 FAI-snail-9 Succinea strigata Fairbanks, Alaska 64°51′N −26.0 −9.9 16.1 −10.8 FAI-snail-10 Succinea strigata Fairbanks, Alaska 64°51′N −26.0 −10.8 15.3 −10.4 FAI-snail-11 Succinea strigata Fairbanks, Alaska 64°51′N −24.9 −10.8 14.1 −10.9 FAI-snail-12 Succinea strigata Fairbanks, Alaska 64°51′N −27.6 −11.0 16.7 −11.1 FAI-snail-13 Succinea strigata Fairbanks, Alaska 64°51′N −24.7 −10.2 14.6 −10.8 FAI-snail-14 Succinea strigata Fairbanks, Alaska 64°51′N −25.9 −10.7 15.1 −10.8 FAI-snail-15 Succinea strigata Fairbanks, Alaska 64°51′N −26.2 −10.0 16.2 −10.3 FAI-snail-16 Succinea strigata Fairbanks, Alaska 64°51′N −26.3 −10.1 16.2 −11.0 FAI-snail-17 Succinea strigata Fairbanks, Alaska 64°51′N −23.0 −9.8 13.2 −11.3 FAI-snail-18 Succinea strigata Fairbanks, Alaska 64°51′N −25.9 −10.4 15.4 −10.6 FAI-snail-19 Succinea strigata Fairbanks, Alaska 64°51′N −25.3 −10.3 15.0 −11.0 FAI-snail-20 Succinea strigata Fairbanks, Alaska 64°51′N −25.9 −11.1 14.9 −10.5 FAI-snail-21 Succinea strigata Fairbanks, Alaska 64°51′N −24.3 −9.7 14.6 −11.3 FAI-snail-22 Succinea strigata Fairbanks, Alaska 64°51′N −25.3 −10.8 14.5 −10.9 FAI-snail-23 Succinea strigata Fairbanks, Alaska 64°51′N −26.5 −10.4 16.2 −10.8 FAI-snail-24 Succinea strigata Fairbanks, Alaska 64°51′N −25.4 −10.6 14.8 −10.8 FAI-snail-25 Succinea strigata Fairbanks, Alaska 64°51′N −10.4 −10.8 FAI-snail-26 Euconulus fulvus Fairbanks, Alaska 64°51′N −11.2 −10.9 FAI-snail-27 Euconulus fulvus Fairbanks, Alaska 64°51′N −10.4 −10.3 FAI-snail-28 Euconulus fulvus Fairbanks, Alaska 64°51′N −10.8 −10.7 FAI-snail-29 Euconulus fulvus Fairbanks, Alaska 64°51′N −10.8 −11.1 FAI-snail-30 Euconulus fulvus Fairbanks, Alaska 64°51′N −10.3 −11.3 FAI-snail-31 Euconulus fulvus Fairbanks, Alaska 64°51′N −10.6 −10.9 FAI-snail-32 Euconulus fulvus Fairbanks, Alaska 64°51′N −10.8 −10.5 FAI-snail-33 Euconulus fulvus Fairbanks, Alaska 64°51′N −10.5 −11.8 FAI-snail-34 Euconulus fulvus Fairbanks, Alaska 64°51′N −10.5 −11.9 FAI-snail-35 Euconulus fulvus Fairbanks, Alaska 64°51′N −11.0 −9.4

intrashell analyses because the goal of this work was to evaluate the after 24 h using the GasBench II connected to an IRMS. Analytical uncer- dominant climatic controlling factors of snails over their annual- tainty was ±0.1‰ forbothcarbonandoxygenisotopes. biannual lifespan. In addition, because analyzed species were consider- ably small (b5 mm maximum length) with quite thin shells, intrashell Stable isotope analysis analyses were not possible. Samples were treated with 3% H2O2 overnight to remove potential organic contaminants. About 150 μg of car- Samples were measured in the stable isotope facility of the Depart- bonate was weighted in a 6 ml ExetainerTM vial that was subsequently ment of Earth and Environmental Sciences, University of Kentucky.

ﬂushed with helium. The carbonate was then converted to CO2 gas by Organic matter samples (snail body tissue, snail feces, plants, fungi adding 0.1 ml of 100% H3PO4 at 25°C. The resulting CO2 was analyzed and soil organic matter) were analyzed in a Costech Elemental Analyzer

Table 2 Stable carbon and oxygen isotope results of modern land snail shells from San Salvador Island (Bahamas) collected during July 2010.

Snail ID Species Locality Latitude Shell δ13C Shell δ18O (PDB) (PDB)

CT-snail-1 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −6.5 −0.2 CT-snail-2 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −5.8 −0.4 CT-snail-3 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −6.3 −0.7 CT-snail-4 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −4.6 +0.1 CT-snail-5 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −5.7 −0.7 CT-snail-6 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −5.5 −0.2 CT-snail-7 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −7.1 +0.8 CT-snail-8 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −6.4 −0.7 CT-snail-9 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −9.0 −1.3 CT-snail-10 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −7.5 −0.7 CT-snail-11 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −7.5 −0.6 CT-snail-12 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −6.1 −0.7 CT-snail-13 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −7.1 −0.5 CT-snail-14 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −4.6 +0.1 CT-snail-15 Succinea barbadensis Cockburn Town, San Salvador 24°60′N −6.9 −0.9 592 Y. Yanes / Quaternary Research 83 (2015) 588–595

Table 3 Stable carbon isotope values of succineid snail feces and potential snail dietary sources in the boreal forest of Fairbanks, Alaska. Each bulk sample represents a mixture of multiple individuals that were homogenized together.

Sample ID Sample type Locality Latitude δ13C (PDB)

FAI-feces-1 Snail feces Fairbanks, Alaska 64°51′N −29.3 FAI-diet-1 Tree leaves Fairbanks, Alaska 64°51′N −28.1 FAI-diet-2 Fungi Fairbanks, Alaska 64°51′N −19.7 FAI-diet-3 Moss Fairbanks, Alaska 64°51′N −34.4 FAI-diet-4 Bulk soil organics Fairbanks, Alaska 64°51′N −27.1

(ESC 4010) connected to a continuous ﬂow isotope ratio mass spec- trometer (IRMS) Finnigan DeltaPLUS XP. Aragonitic shells were measured in a GasBench II connected to the same IRMS. All stable isotope results are reported in δ notation relative to the international standard Pee Dee Belemnite (PDB) for both organic matter and carbonate samples. The δ values are deﬁned as: hi δ13 δ18 ¼ = – ðÞ‰ Cor O Rsample Rstandard 1 1000 where R¼13C=12Cor18O=16O:

Results

Fairbanks

Two species of land snails were found during a field survey in August 2010 in a boreal forest of Fairbanks (Table 1): (1) Succinea strigata (n = 25), with a maximum shell length of ~5 mm (Fig. 1C), and (2) E. fulvus (n = 10), a microsnail with maximum shell length of ~2 mm. Specimens were found alive attached to fallen logs on the forest floor. Despite their difference is size, both species showed statistically equivalent (t-test, p N 0.05) carbon and oxygen isotope values in their shells, and therefore, they are treated collectively. Shell δ13C values ranged from −11.2‰ to −9.7‰ (Figs. 3A–B), with an average value of −10.4 ± 0.4‰ (n = 35). Bulk body tissue was measured for 20 Succinea specimens. Body δ13C values ranged from −27.6‰ to −23.0‰ (Fig. 3B), averaging −25.5 ± 1.0‰ (n = 20). Shell and body δ13Cvaluescorrelatedweakly(- Fig. 3B). While shells showed a reduced range in δ13C values (Δ13C= 1.4‰), body tissues exhibited significant variations among specimens (Δ13C=4.6‰). Shells were, on average, 15‰ 13C-enriched with respect to body tissue (Fig. 4A). Potential food resources for land snails in the studied habitat included C3 vascular plants, moss, fungi and decayed soil organic matter. One sample of each potential carbon source was mea- δ13 − ‰ − ‰ sured and displayed respective C values of 28.1 , 34.4 , Figure 3. Stable isotope results of land snails from Fairbanks. (A) Oxygen and carbon stable −19.7‰ and −27.1‰ (Table 3; Fig. 4A). Finally, a homogenized sample isotope values of modern land snail shells. (B) Relationship between carbon stable isotope of snail feces from multiple Succinea specimens showed a δ13Cvalueof values of succineid land snail shell and body tissue from Fairbanks. (C) Carbon and oxygen −29.3‰ (Fig. 4A). Snail body tissue was ~3.8‰ higher in 13C than snail stable isotope values of land snail shells from Fairbanks and San Salvador Island. feces (Table 3). The δ18O values of the snail shells from Fairbanks ranged from higher in δ18O than local rainfall (Fig. 4B). Snail shell δ18O values from −11.9‰ to −9.4‰ (Figs. 3A–B), averaging −10.8 ± 0.4‰ (n = 35). the Bahamas were, on average, ~10‰ higher (in PDB scale) than shells On average, shell δ18O values are ~4.2‰ 18O-enriched with respect to from Fairbanks (Fig. 4B). Similarly, annual rainfall δ18O values from the summer rainwater from the region (Fig. 4B). Bahamas were, on average, ~10‰ higher (in SMOW scale) than summer rain from Fairbanks (Fig. 4B). San Salvador, Bahamas Discussion For comparison, several specimens of S. barbadensis collected in July of 2010 in Cockburn Town, San Salvador (Bahamas), located at the lati- Carbon stable isotopes tude of 24°N (Figs. 1A, D–E), were analyzed isotopically (Table 2). The shell δ13C showed an average value of −6.4 ± 1.0‰ (n = 10), ranging Stott (2002) and Metref et al. (2003) showed that snails primarily from −9.0‰ to −4.6‰. Thus, shells from Bahamas were ~4‰ higher in incorporate carbon isotopes from the plant diet in their tissues, whereas 13 C than counterparts from Fairbanks. other carbon sources such as atmospheric CO2 and carbonate ingestion Shell δ18O values ranged from +0.8‰ to −1.3‰ (Fig. 3C), with an had a negligible influence. Thus, the carbon isotope composition of snail average value of −0.4 ± 0.5‰ (n = 10). Bahamian shells were ~4.9‰ tissues has been used as a proxy for paleovegetation at low-to-mid Y. Yanes / Quaternary Research 83 (2015) 588–595 593

Figure 5. Calculated curves of land snail shell δ13C values as a function of body δ13Cvalues, which depict snail's diet, based on the steady-state ﬂux balance model by Balakrishnan and Yapp (2004). Calculations were made assuming that calciﬁcation occurred at average air temperatures of 13°C.

rates among sympatric specimens. The snail evaporative steady-state ﬂux balance model developed by Balakrishnan and Yapp (2004) was used to evaluate whether or not snails from Fairbanks experienced no- ticeably different metabolic rates. The model relates the amount and isotopic composition of food resources, bicarbonate in the snail hemo-

lymph, and the diffusive flux of respired CO2.Theflux of dissolved bicarbonate output from the hemolymph (fo)relativetotheflux of diet (fin) is called ϕ (=fo / fin) and varies with metabolic rate (Balakrishnan and Figure 4. (A) Carbon stable isotope values of land snail tissues and potential snail food re- Yapp, 2004). Model calculations were constrained by the measured 13 sources from Fairbanks, Alaska. Triangles depict results from snail tissues, including feces δ C values of shell and body tissue and the mean air temperature dur- (open symbol), body tissue (gray symbol) and shell (black symbol). Circles depict plant ing snail active period, assumed to be ~13°C. The model, in combination samples (moss = black symbol; tree leaves = gray symbol). Open "quadrat depicts soil with the δ13C values of snails from Fairbanks, indicates that specimens organic matter and open diamond depicts fungi. Note that only samples of snail shell ϕ and body included multiple individuals whereas the remaining samples depict a single varied in metabolic rate, with ranging from 0.00 to 0.20 (Fig. 5). This 13 analysis. (B) Comparison of the oxygen stable isotope values of land snail shells suggests that variations in shell δ C values may be in part explained (diamonds) and rainfall (circles) between Fairbanks (filled symbols) and San Salvador by differing metabolic rates among specimens. This finding differs from Island (open symbols). some published results, which did not observe significant variations in model-derived metabolic rates among Sphincterochila candidissima spec- latitude sites. While studies in carbonate-rich areas suggest that snails imens (Yanes et al., 2013a)andTheba geminata individuals (Yanes et al., may incorporate carbon isotopes derived from limestone ingestion, 2013b) both from carbonate-rich areas from mid- and low-latitude sites, which complicates plant ecology inferences (Goodfriend and Hood, respectively. Accordingly, potential variations in snail metabolic rates 1983; Goodfriend, 1987; Yanes et al., 2008, 2012, 2013a), snails that oc- appear to be species-dependent and require additional examination). cupy carbonate-poor areas (e.g., forests with acidic soils), may offer Succinea specimens from the tropical island of San Salvador more accurate information about local vegetation cover. In the case of (Bahamas) showed a wider range of shell δ13Cvalues,andwere,on the studied boreal forest here, soil samples did not contain calcium car- average, ~4‰ higher than those from Fairbanks (Fig. 3C). The higher bonate, so the influence of carbonate ingestion in snail shells is assumed variability in shell δ13C values of specimens from San Salvador is ex- to be negligible. However, snails may utilize other carbon sources in ad- plained by the higher variation in δ13C values of potential carbon dition to vascular plants. In fact, in woodland areas, snails consume, sources in the island, which includes the presence of both C3 and C4 apart from C3 plants, other food resources including moss, fungi, algae, plants, and carbonate-rich marine sediments (Baldini et al., 2007; tree sap, and decayed organic matter (Speiser, 2001), which, in turn, Yanes and Romanek, 2013). The results from the present work suggest may differ significantly in carbon isotope values (Fig. 4A). Snail body tis- that even when individuals follow variable metabolic rates (like those sue is enriched in 13C with respect to diet by ~1‰ (DeNiro and Epstein, from Fairbanks), higher dispersion in δ13C values is expected at 1978; Stott, 2002; Metref et al., 2003). Subtraction of 1‰ from the aver- carbonate-rich localities where multiple photosynthetic pathways co- age δ13C of the body tissue of the Fairbanks snails yields a value of exist. This study also suggests that snails from boreal forests, which −26.5 ± 1.0‰.InFig. 4A it is illustrated that the average snail body grow and reproduce in a narrow temporal window during the warmest δ13C value plots close to the value of fresh tree leaves (−28.1‰) and (ice-free) months, may experience higher variability of metabolic rates the soil organic matter sample (−27.1‰), suggesting that the mea- than snails from mid-to-low latitudes, with longer active periods sured Succinea specimens from Fairbanks included significant propor- throughout the year. tions of living and decayed C3 plant matter in their diet, probably with some assimilation of fungi (−19.7‰). It seems less likely that Succinea Oxygen stable isotopes included moss (−34.4‰) in its diet. An additional complication in interpreting the carbon isotope com- Snail shells from Fairbanks exhibited a rather narrow range of δ18O position of land snail shells is the potential variations in metabolic values, averaging −10.8 ± 0.4‰ (Figs. 3A–B). This is probably one of 594 Y. Yanes / Quaternary Research 83 (2015) 588–595 the most negative oxygen isotope values for land snail shells reported in shells year-round in San Salvador, shells should have been deposited the published literature. Even though specimens of Succinea and at times when relative humidity was ~86%, on average (Fig. 6). This cal- Euconulus exhibited statistically comparable δ18O values, the microsnail culated value of relative humidity during calcification overlaps with Euconulus displayed a larger dispersion than Succinea (Fig. 3A). This predicted values by Yanes and Romanek (2013) for modern Cerionidae larger dispersion suggests that the microsnail may have experienced and Chondropomidae snails collected at different localities within the higher evaporation rates than Succinea specimens. The shell δ18O values same island. Hence, different snail species from the Bahamas seem to of snails from Fairbanks were examined using the evaporative steady- have grown under comparable RH values. The results reported here state flux balance-mixing model published by Balakrishnan and Yapp suggest that succineid snail shells from the Bahamas were deposited (2004). The model links the amount and isotopic composition of exter- at drier conditions than specimens from Fairbanks. nal liquid water, liquid water from the snail hemolymph, the diffusive Despite the complexity associated with multiple (rather than one) flux of water from the hemolymph by evaporation, and the temperature climatic variables controlling snail shell δ18O values, meaningful paleo- dependent oxygen isotope fractionation between water and aragonite climatic inferences can be deduced from fossil snails, especially at coarse (Grossman and Ku, 1986). This model considers that δ18Ovaluesof temporal/spatial scales. This study shows that at a coarse latitudinal water and water vapor, relative humidity and temperature are the scale, shell δ18O values seem to be predominantly a function of input most important variables that control shell δ18O values (Balakrishnan rainfall δ18O values. Snail shells recovered from Quaternary outcrops and Yapp, 2004). The model considers the flux of liquid water output in Alaska will provide valuable summer season paleoclimatic data, 18 from the hemolymph (fo) relative to the flux of liquid water imbibed especially paleorainfall δ O values and paleohumidity. (fin), a ratio called θ (=fo / fin). Model calculations assume that ambient water vapor is in isotope equilibrium with the imbibed liquid water and Conclusions that water is lost by evaporation (Balakrishnan and Yapp, 2004). Model outputs are constrained by the measured shell carbonate δ18O values, Land snail shells from high latitudes are useful proxies for summer the summer temperature in Fairbanks (~13°C), and the δ18O values of climatic conditions. Land snails from Fairbanks, interior Alaska, primar- the summer rainfall (−15‰ vs. SMOW). If land snails from Fairbanks, ily recorded the carbon isotope values of the snail's diet in their shell δ18 − ‰ with a measured shell Ovalueof 10.8 , grew their shells at and body tissue, which included living and decayed C3 plant matter times when temperatures were about 13°C and rainwater δ18O values and some assimilation of fungi. A flux-balance mixing model suggested were around −15‰, the model predicts that relative humidity condi- that measured individuals exhibited significantly different metabolic tions during calcification were near 93% (Fig. 6). This relatively high rates, in contrast to some published data from mid to low latitudes. value for RH is indeed reached during the summer months in Fairbanks, Thus, variations in shell δ13C values among individuals of the same spe- indicating that Succinea and Euconulus specimens from the study area cies may result from variable metabolic rates rather than from variable grow at notably moist conditions during the summer months. diets. The oxygen isotope composition of snails from Fairbanks Succinea shells from the tropical island of San Salvador (Bahamas) (−10.8 ± 0.5‰ vs PDB) is likely the most negative value reported for displayed an average δ18O value of −0.4‰, that is, 10.4‰ higher than land snails. A flux-balance mixing model indicates that analyzed snails snails from Fairbanks (Fig. 3C). The average annual rain δ18O value in from Fairbanks primarily deposited their shell during the summer sea- San Salvador is −4.5‰(SMOW) and the annual mean temperature is son, when relative humidity was ~93%. A comparison between 24°C (Baldini et al., 2007; Yanes and Romanek, 2013). Annual rainwater succineid specimens from Alaska (64°N) and the tropical island of San δ18O values from San Salvador are ~10‰ higher than summer rainwater Salvador (24°N) suggests that the local δ18O values of the rainfall during values from Fairbanks (Fig. 3C). Thus, the isotopic offset between shells snail active periods are probably the dominant control of shell δ18O from Fairbanks and San Salvador is equivalent to the observed offset values at a coarse latitudinal scale. While at the microhabitat scale, the between respective rainfall isotopic values. Using the model by shell δ18O values might be complex to understand due to the high Balakrishnan and Yapp (2004) and assuming that snails grew their number of environmental factors operating collectively, the shell δ18O values at rough latitudinal scales seem to be a meaningful proxy for rainfall δ18O.

Acknowledgments

This study was in part supported by the Spanish Ministry of Economía y Competitividad (MEC) grant CGL2011-29898/BTE to Y.Y. Special thanks go to Chris Romanek who let the author personally run all samples used in this study in the Stable Isotope Facility of the University of Kentucky. The author is grateful to QR editors Robert Booth, Alan Gillespie and Michael O'Neal; and to reviewers Crayton Yapp, Giovanni Zanchetta and André Colonese for their critical and de- tailed reviews, which signiﬁcantly improved the clarity and quality of this manuscript.

References

Ansart, A., Vernon, P., Daguzan, J., 2002. Elements of cold hardiness in a littoral population of the land snail Helix aspersa (Gastropoda: Pulmonata). Journal of Comparative Phys- iology B 172, 619–625. Balakrishnan, M., Yapp, C.J., 2004. Flux balance model for the oxygen and carbon isotope compositions of land snail shells. Geochimica et Cosmochimica Acta 68, 2007–2024. Balakrishnan, M., Yapp, C.J., Theler, J.L., Carter, B.J., Wyckoff, D.G., 2005a. Environmental δ18 Figure 6. Calculated curves of land snail shell O values as a function of relative humidity signiﬁcance of 13C/12Cand18O/16O ratios of modern land-snail shells from the south- ﬂ (RH) using the evaporative steady-state ux balance-mixing model by Balakrishnan and ern Great Plains of North America. Quaternary Research 63, 15–30. Yapp (2004). Curves were calculated for two temperatures (24°C and 13°C) and rainfall Balakrishnan, M., Yapp, C.J., Meltzer, D.J., Theler, J.L., 2005b. Paleoenvironment of the 18 δ O values (−4.5‰ and −15.0‰). Arrows illustrate that, on average, succineid shells Folsom archaeological site, New Mexico, USA, approximately 10,500 14CyrB.P.asin- from San Salvador grew their shells at RH of ~86% while succineid shells from Fairbanks ferred from the stable isotope composition of fossil land snail shells. Quaternary were deposited at times when RH was ~93%. Research 63, 31–44. Y. Yanes / Quaternary Research 83 (2015) 588–595 595

Baldini, L.M., Walker, S.E., Bruce, R., Baldini, J.U.L., Crowe, D.E., 2007. Isotope ecology of the Nicolai, A., Filser, J., Briand, V., Charrier, M., 2010. Seasonally contrasting life-history modern land snails Cerion, San Salvador, Bahamas: preliminary advances toward strategies in the land snail Cornu aspersum: physiological and ecological implications. establishing a low-latitude island palaeoenvironmental proxy. Palaios 22, 174–187. Canadian Journal of Zoology 88, 995–1002. Colonese, A.C., Zanchetta, G., Fallick, A.E., Martini, F., Manganelli, G., Lo Vetro, D., 2007. Nicolai, A., Filser, J., Lenz, R., Bertrand, C., Charrier, M., 2012. Quantitative assessment of Stable isotope composition of late glacial land snail shells from Grotta del Romito hemolymph metabolites in two physiological states and two populations of the (Southern Italy): paleoclimatic implications. Palaeogeography Palaeoclimatology Pa- land snail Helix pomatia. Physiological and Biochemical Zoology 85, 274–284. laeoecology 254, 550–560. Örstan, A., 2010. Reproductive biology and annual population cycle of Oxyloma retusum Colonese, A.C., Zanchetta, G., Fallick, A.E., Martini, F., Manganelli, G., Russell, D., 2010a. (Pulmonata: Succineidae). American Malacological Bulletin 28, 113–120. Stable isotope composition of Helix ligata (Müller, 1774) from Late Pleistocene–Holocene Pearce, T.A., Örstan, A., 2006. Terrestrial Gastropoda. In: Sturm, C.F., Pearce, T.A., Valdés, A. archaeological record from Grotta della Serratura (Southern Italy): paleoclimatic (Eds.), The Mollusks: A Guide to Their Study, Collection, and Preservation. American implications. Global and Planetary Change 71, 249–257. Malacological Society. Colonese, A.C., Zanchetta, G., Dotsika, E., Drysdale, R.N., Fallick, A.E., Grifoni-Cremonesi, R., Pigati, J.S., Rech, J.A., Nekola, J.C., 2010. Radiocarbon dating of small terrestrial gastropod Manganelli, G., 2010b. Early–middle Holocene land snail shell stable isotope record shells in North America. Quaternary Geochronology 5, 519–532. from Grotta di Latronico 3 (southern Italy). Journal of Quaternary Science 25, Pigati, J.S., McGeehin, J.P., Muhs, D.R., Bettis, E.A., 2013. Radiocarbon dating late Quaterna- 1347–1359. ry loess deposits using small terrestrial gastropod shells. Quaternary Science Reviews Colonese, A.C., Zanchetta, G., Russell, D., Fallick, A., Manganelli, G., Lo Vetro, D., Martini, F., 76, 114–128. di Giuseppe, Z., 2011. Stable isotope composition of Late Pleistocene–Holocene Speiser, B., 2001. Food and feeding behavior. In: Barker, G.M. (Ed.), The Biology of Eobania vermiculata shells (Müller, 1774) (Pulmonata, Stylommatophora) from the Terrestrial Mollusk. CABI, pp. 259–288. Central Mediterranean basin: data from Grotta d'Oriente (Favignana, Sicily). Stott, L.D., 2002. The influence of diet on the δ13C of shell carbon in the pulmonate snail Quaternary International 244, 76–87. Helix aspersa. Earth and Planetary Science Letters 195, 249–259. Colonese, A.C., Zanchetta, G., Fallick, A.E., Manganelli, G., Saña, M., Alcade, G., Nebot, J., Yanes, Y., Romanek, C.S., 2013. Quaternary interglacial environmental stability in San 2013a. Holocene snail shell isotopic record of millennial scale hydrological conditions Salvador Island (Bahamas): a land snail isotopic approach. Palaeogeography in western Mediterranean: data from Bauma del Serrat del Pont (NE Iberian Palaeoclimatology Palaeoecology 369, 28–40. Peninsula). Quaternary International 303, 43–53. Yanes, Y., Delgado, A., Castillo, C., Alonso, M.R., Ibáñez, M., De la Nuez, J., Kowalewski, M., Colonese, A.C., Zanchetta, G., Perlès, C., Drysdale, R.N., Manganelli, G., Baneschi, I., Dotsika, 2008. Stable isotope (δ18O, δ13C, and δD) signatures of recent terrestrial communities E., Valladas, H., 2013b. Deciphering late Quaternary land snail shell δ18O and δ13C from a low-latitude, oceanic setting: endemic land snails, plants, rain, and carbonate from Franchthi Cave (Argolid, Greece). Quaternary Research 80, 66–75. sediments from the eastern Canary Islands. Chemical Geology 249, 377–392. DeNiro, M.J., Epstein, S., 1978. Influence of diet on the distribution of carbon isotopes in Yanes, Y., Romanek, C.S., Delgado, A., Brant, H.A., Noakes, J.E., Alonso, M.R., Ibáñez, M., animals. Geochimica et Cosmochimica Acta 42, 495–506. 2009. Oxygen and carbon stable isotopes of modern land snail shells as environmen- Gomot de Vaufleury, A., 2001. Regulation of growth and reproduction. In: Barker, G.M. tal indicators from a low-latitude oceanic island. Geochimica et Cosmochimica Acta (Ed.), The Biology of Terrestrial Mollusk. CABI, pp. 331–356. 73, 4077–4099. Goodfriend, G.A., 1987. Radiocarbon age anomalies in shell carbonate of land snails from Yanes, Y., Yapp, C.J., Ibáñez, M., Alonso, M.R., De la Nuez, J., Quesada, M.L., Castillo, C., semi-arid areas. Radiocarbon 29, 159–167. Delgado, A., 2011a. Pleistocene–Holocene environmental change in the Canary Archi- Goodfriend, G.A., 1992. The use of land snail shells in paleoenvironmental reconstruction. pelago as inferred from stable isotopes of land snail shells. Quaternary Research 65, Quaternary Science Reviews 11, 665–685. 658–669. Goodfriend, G.A., 1999. Terrestrial stable isotope records of late Quaternary paleoclimates Yanes, Y., Romanek, C.S., Molina, F., Cámara, J.A., Delgado, A., 2011b. Holocene in the eastern Mediterranean region. Quaternary Science Reviews 18, 501–513. Paleoenvironment (7,200–4,000 cal. years BP) of the Los Castillejos Archaeological Goodfriend, G.A., Ellis, G.L., 2000. Stable carbon isotope record of middle to late Holocene site (Southern Spain) as inferred from stable isotopes of land snail shells. Quaternary climate changes from land snail shells at Hinds Cave. Texas. Quaternary International International 244, 67–75. 67, 47–60. Yanes, Y., Gutiérrez-Zugasti, I., Delgado, A., 2012. Late glacial–early Holocene transition in Goodfriend, G.A., Ellis, G.L., 2002. Stable carbon and oxygen isotope variations in modern northern Spain deduced from terrestrial gastropod shelly accumulations. Quaternary Rabdotus land snail shells in the southern Great Plains, USA, and their relation to Research 78, 373–385. environment. Geochimica et Cosmochimica Acta 66, 1987–2002. Yanes, Y., García-Alix, A., Asta, M.P., Ibáñez, M., Alonso, M.R., Delgado, A., 2013a. Late Goodfriend, G.A., Hood, D.G., 1983. Carbon isotope analysis of land snail shells: Pleistocene–Holocene environmental conditions in Lanzarote (Canary Islands) in- implications for carbon sources and radiocarbon dating. Radiocarbon 25, 810–830. ferred from calcitic and aragonitic land snail shells and bird bones. Palaeogeography Grossman, E.L., Ku, T.L., 1986. Oxygen and carbon isotope fractionation in biogenic Palaeoclimatology Palaeoecology 378, 91–102. aragonite. Chemical Geology 59, 59–74. Yanes, Y., Gómez-Puche, M., Esquembre-Bebia, M.A., Fernández-López-de-Pablo, J., 2013b. Kehrwald, N.M., McCoy, W.D., Thibeault, J., Burns, S.J., Oches, E.A., 2010. Paleoclimatic im- Younger Dryas – early Holocene transition in southeastern Iberian Peninsula: insights plications of the spatial patterns of modern and LGM European land-snail shell δ18O. from land snail shell middens. Journal of Quaternary Science 28, 777–788. Quaternary Research 74, 166–176. Yanes, Y., Izeta, A.D., Cattaneo, R., Costa, T., Gordillo, S., 2014. Holocene (~4.5-1.7 cal. kyr Lécolle, P., 1985. The oxygen isotope composition of land snail shells as a climatic indica- BP) paleoenvironmental conditions in central Argentina inferred from entire-shell tor: applications to hydrogeology and paleoclimatology. Chemical Geology 58, and intra-shell stable isotope composition of terrestrial gastropods. The Holocene 157–181. 24, 1193–1205. Metref, S., Rousseau, D.D., Bentaleb, I., Labonne, M., Vianey-Liaud, M., 2003. Study of the Yapp, C.J., 1979. Oxygen and carbon isotope measurements of land snail shell carbonates. diet effect on δ13C of shell carbonate of the land snail Helix aspersa in experimental Geochimica et Cosmochimica Acta 43, 629–635. conditions. Earth and Planetary Science Letters 211, 381–393. Zanchetta, G., Leone, G., Fallick, A.E., Bonadonna, F.P., 2005. Oxygen isotope composition Moine, O., Rousseau, D.D., Antoine, P., 2005. Terrestrial molluscan records of Weichselian of living land snail shells: data from Italy. Palaeogeography Palaeoclimatology Lower to Middle Pleniglacial climatic changes from the Nussloch loess series Palaeoecology 223, 20–33. (Rhine Valley, Germany): the impact of local factors. Boreas 34, 363–380. Zang, N., Yamada, K., Suzuki, N., Yoshida, N., 2014. Factors controlling shell carbon isotopic Moine, O., Rousseau, D.D., Antoine, P., 2008. The impact of Dansgaard-Oeschger cycles on composition of land snail Acusta despecta sieboldiana estimated from lab culturing the loessic environment and malacofauna of Nussloch (Germany) during the Upper experiment. Biogeosciences 11, 6555–6590. Weichselian. Quaternary Research 70, 91–104.